YEHAND NOYES
2978
VOL.27
Formation and Decomposition of Positive Iodine Esters of Carboxylic Acids SI-JUNGYEH AND RICHARD M. N o m s Department of Chemistry, University of Oregon, Eugene, Oregon Received May 16,1989 Silver propionate and iodine in the molar ratio 3 :2 react quantitatively to form ( CZHSCOO)J. I n saturated hydrocarbon Eolvent, thia compound decomposes to ethyl iodide and unidentified other products with a half-life of several days in the dark at room temperature. The reaction is accelerated by light. Excess iodine causes a more rapid decomposition that is also accelerated by light and that may involve intermediate formation of CyH&OOI. The ethyl iodide formed in this more rapid decomposition contains iodine that haa been isotopically equilibrated with the added iodine, and even a large excese of elementary iod&e fails to produce ethyl iodide quantitatively from the propionyl groups of the original iodine(II1) ester. Concentrations of ethyl iodide and of positive iodine undergo peculiar coupled fluctuations during the initial stages of the reaction, and ethyl iodide is consumed aa well aa produced during certain stages of the reaction. The system is too complex to provide information about geminate recombination during the decomposition of propionyl hypoiodite.
Some organic molecules dissociate in solution to form stable molecules and pairs of radical fragments that can either undergo geminate reaction or diffuse apart. The products of reaction under various conditions can be used to determine the probability that these fragments will become separated by diffusion. Compounds studied in this way include the azonitriles,’ R-N=N-R, and the aliphatic acyl peroxides,*J (RC00)z. The acyl hypoiodites, RCOOI, should decompose similarly to yield alkyl iodide and carbon dioxide. If such a decomposition were carried out in a solution containing isotopically labeled elementary iodine, the alkyl iodide formed by geminate recombination of radical fragments would be unlabeled, while radicals that escaped by diffusion would form labeled alkyl iodide. Although alkyl hypoiodites have not been isolated, they have been postulated as intermediates during the reaction of iodine with the silver salts of carboxylic acids. The literature on this reaction and the mechanistic proposals have been reviewed fairly recently.‘ We report here some additional studies of the reaction that provide new information of mechanistic value even though they provide little insight on the specific problem of geminate recombination. Results
Formation of Positive Iodine Esters.-Equimolar amounts of solid silver propionate and iodine in hexane solvent could be expected to react according to the equation GHrCOOAg
+ 1, +CZHsCOOI + AgI
(1)
~~
(1) 0. 8. Hammond, C. H. 8. Wu,0. D. Trapp, J. Warkentin, and 6394 (1980). (2) M. Bswaro, J . Polymer Sn’., 16, 387 (1955). (3) D. F. DeTar and R. C. Lamb, J . Am. Chum. Soo., 81, 122 (1959). (4) J. Kleinberg, Chum. Raw., 40, 381 (1947). (6) R. a. Johnson and R. K. Ingham, ibid., 66,219 (1968). (8) C. V. Wilaon, “Organic Reaotions,” Vd. IX, John Wiley and SOM,New York, 1957, p. 332.
R. T. Keya, J. Am. Chum. Soc., 88,
However, reactants in these proportions never produced solutions containing significant amounts of species capable of oxidizing aqueous iodide. As is shown below, propionyl hypoiodite decomposes rapidly at room temperature and probably was destroyed as rapidly as it was formed. If the molar ratio of silver propionate to iodine is increased from 1.0 to 1.5, the observations of Oldham and Ubbelohde’ suggest the reaction 3C2HsCOOAg
+ 212 --+ (CzHsC00)sI + 3AgI
(2)
Reaction under these conditions was somewhat accelerated by light and took place quantitatively in sixty to one hundred and fifty minutes at room temperature. The stoichiometry was confirmed to within about 201, by radiochemical procedures (which showed that one quarter of the initial iodine remained in the now colorless hexane solution) and by hydrolysis of the iodine(II1) tripropionate to iodate according to the equation 5(C2HsCOO)sI
+ 9HzO + 3HIOa
+ Is + 15CzHsCOOH
(3)
The existence of the iodate was confirmed both by reducing with sulfite and titrating the resulting iodide with silver nitrate and also by reducing with iodide and titrating with thiosulfate. The iodidethiosulfate procedure was used in the subsequent studies of the decomposition reaction. Since each mole of iodate requires six equivalents of thiosulfate for titration, and since the iodine formed in equation 3 remains in the hexane layer, each ten oxidizing equivalents of iodine(II1) initially present produce nine oxidizing equivalents (as iodate) in the aqueous layer and one oxidizing equivalent (as elementary iodine) in the hexane layer. Decomposition of Positive Iodine Esters.-Dry hexane solutions of iodine (111) tripropionate decomposed slowly, and the reaction could be followed conveniently by measuring the oxidizing titer of (7) J. W. H. Oldhsm and A. R. Ubbelohde, J. Chem. Soc., 388 (1941).
SEPTEMBER, 1962
POSITIVE IODINE ESTERSOF CARBOXYLIC ACIDS
the iodate formed by treating an aliquot with water as indicated in equation 3. The rate of decomposition in the dark was about one eighth that under normal laboratory illumination. I n the illuminated solution, the iodine concentration remained very low, and the decomposition produced ethyl iodide. In the dark, the ethyl iodide concentration remained constant, and the decomposition produced iodine. Since the reaction was only followed about 30% of the way to completion, the kinetics were not established with certainty. However, the data seemed to be fitted best by the assumption of first order or less dependence on iodine(II1) tripropionate. The apparent first order rate constant in the dark at 25' was 1.34 X sec.-l, corresponding to a half-life of about six days. The decomposition reaction is obviously complex. Iodine atoms can accelerate at least one possible path and are consumed in the process, Effects of Added Iodine.-If excess elementary iodine were added to a solution of iodine(II1) tripropionate, the data of Oldham and Ubbelohde' suggest thqt the equilibrium (CzHIC00)aI
+ Io
3CpHsCOOI
(4)
should be established. An analogy with equation 3 predicts that propionyl hypoiodite should also disproportionate with water according to the equation 5CzHsCOOI
+ 3Hz0 +HI03 + 212 + 5C2HrCOOH ( 5 )
Of the ten oxidizing equivalents of iodine(1) on the left side of equation 5, six will be present in the aqueous layer as iodate and four will be present in the hexane layer as additional iodine. Hence equation 5 predicts a ratio of 1.5 for oxidizing equivalents formed in aqueous and hexane layers compared to the ratio of 9.0 predicted by equation 3 and observed in the absence of excess iodine. Our experiments failed to confirm equilibrium 4. When excess iodine was added to solutions of ( C Z H ~ C O Othe ) ~ ~iodine , color faded and the rate of loss of oxidizing titer greatly accelerated. The reaction was fastest in solutions exposed to light, but even in the dark the half time for decomposition was reduced from a few days to a few hours. When water was added, the ratio of equivalents of aqueous oxidizing titer per equivalent of additional iodine in hexane fell from nine but usually did not go much below four. Only occasionally and with large excess of iodine did the ratio approach the 1.5 predicted by equation 5. If concentrations of individual species were computed by assuming that equations 3 and 5 described the effects of water and that the only oxidizing species present were those in equilibrium 4, the calculated equilibrium constant decreased as a run progressed. This behavior would be expected if the rate of decomposition of CzH~COOI
2979
l
P: 0.014
w
2 0.012 ii
g 0.010 E
ij
0.004 1 i
0 . 0 1 , , 2
4
,
,
6
8
,
10 TIME,HR.
:
,
12
14
HIO, ,
16
f
18
Fig. 1.-Concentrations of species following dark addition of iodine to solution of (CzHsC0O)J. Molar ratio of added iodine to initial (C2H&00)31, 4.80: 1 Concentration of HIOs = moles of this substance extracted with water per liter of hexane solution so treated
were comparable to or greater than the rate of equation 4 from right to left so that the equilibrium could not be established. A question about the analytical method is raised in the Discussion, but we do not see how the data can be accommodated to the establishment of equation 4 as a true equilibrium. The situation is even more complex than is suggested by the above discussion. Figure 1 illustrates the subsequent behavior when unlabeled iodine solution was added in the dark to a solution of iodine(II1) tripropionate that had been prepared with labeled Ilal. Concentrations indicated at zero time are the ones that would have been observed if the dilution had been made with the same volume of hexane instead of with the iodine solution. If all of the positive iodine were initially in the +3 state, the molar ratio of added iodine to initial (CzHsCOO)J was 4.80 to 1. As Fig. 1 shows, immediately after the iodine was added, the titer of water extractable €€IO3 fell rapidly, and the concentration of ethyl iodide rose from 0.00561 M to 0.0111 M . This concentration is more than the 0.01040 M expected if only the iodine in the initial (C2H&!OO)31 were converted to ethyl iodide. The additional ethyl iodide was almost free of P a l ; the iodine from which it was formed must have been isotopically equilibrated with the added iodine. This initial reaction was followed by another that consumed ethyl iodide and increased the titer of water extractable HIOa. When this reaction was over, the ethyl iodide concentration increased again while the oxidizing titer fell with a half life of about five hours (compared to six days in the absence of added iodine). The final concentration of ethyl iodide was only about two thirds of the theoretical maximum of 0.0200 M based on ethyl groups in the original propionate.
2980
VOL.27
YEH A N D NOYEB
I t is tempting to decide that the apparent is very hard to see how the total oxidizing power of maxima and minima in Fig. 1 are artifacts arising the hexane solution could increase during any stage from analytical errors and that they only happen of the reaction. Our experiments determined to coincide in time. However, the procedures water-extractable oxidizing material capable of were carefully verified in tests of apparent stoichiom- oxidizing iodide. When the only oxidizing species etry, and the analyses at each time in Fig. 1 added in the hexane was iodine(II1) tripropionahe, we to the same total iodine content of the system with demonstrated that HI03 was the only oxidizing a total spread of 2.3%. Moreover, the same pat- agent, but equation 3 shows that 10% af the potern was observed repeatedly. When the added tential oxidizing power was not extracted by wakr. excess iodine did not greatly exceed the iodine(II1) If the hexane layer contains propionyl hypoiodite, ester, the maxima and minima were ripples that 40% of its potential oxidizing power is not exscarcely seemed real except that HIo3 and ethyl tracted by water. iodide always changed concentrations in opposite Hence, the water-extractable oxidizing power directions. When the added iodine was in large would be increased by any process that formed excess, the fluctuations were large, as illustrated in iodine(II1) ester at the expense of iodine(1) ester. Fig. 1. Sometimes the concentrations showed two However, it is hard to interpret the data in this positions of relative maximum for HI02 instead way because the increases in apparent HI03 occur of the one shown here. Reproducibility of sup- at precisely the times when the consumption of posedly duplicate runs was far from satisfactory. ethyl iodide is presumably associated with conEffects of Added Ethyl Iodide.-Ethyl iodide version of (111) ester to (I) ester. was expected to be an unreactive product in this An alternative possibility is that propionyl hyposystem. Results like those in Fig. 1 suggest that iodite is water extractable as such. It would the role of ethyl iodide is far from passive. hl- presumably oxidize iodide by the reaction though its concentration usually increased during a CzHsCOOI I- +CzH6COO- -P 11 (8) run, it was consumed during some runs in the dark at low iodine concentration. One solution of iodine and all of the oxidizing power of the iodine(1) ester (111) tripropionate to which about 0.01 M ethyl would then become water-extractable. Although iodide was added decomposed in the dark several this interpretation could explain many of our obtimes faster than did one to which no addition had servations, a few increases in water-extractable oxidizing power were too great to be explained in been made. this way We see no alternative explanations without proposing additional intermediates such Discussion as iodine(I1) esters. Both the formation and decomposition reactions These experiments were undertaken to obtain are photochemically accelerated. Radical proc- information about the dissociation esses can obviously make major contributions to GH6COOI +CZHS.+ COz I. (9) these reactions, but we cannot exclude the possibility of parallel nonradical mechanisms. Previous ~ o r k l -suggests ~ that a significant fracDuring all of our initial thinking, ethyl iodide tion of such dissociations should lead to geminate was assumed to be an inert product in this system. recombination. However, Fig. 1 shows that when Observations like Fig. 1 require revision of this a solution of isotopically labeled (C*H&00)31 assumption. The reaction and ethyl iodide is treated with an excess of unlabeled elementary iodine, little additional labeled C2H5* + CZHd +CZHe + CrHa + 1. (6) ethyl iodide is produced. This observation can be is approximately thermoneutral and may well have interpreted to indicate that reaction 9 is followed a rather low activation energy. The photochem- by very little geminate recombination. Alternaical observations demonstrate that iodine atoms tive interpretations are that the destruction of can accelerate the decomposition of iodine(II1) labeled ethyl iodide by reaction 6 approximately tripropionate. A possible reaction is balanced geminate recombination, that CtH&OOI 1. + (C2HjCOO)aI +2C2HsCOOI + CzHs' + COn ( 7 ) exchanged rapidly with elementary iodine, or that decomposition did not occur by reaction 9 but by Reactions 6 and 7 provide a chain process by some other process such as which the conversion of iodine(II1) to iodine(1) CZHbCOOI* + 1. +C2HJ + COz I** (10) ester can induce decomposition of ethyl iodide. The system is obviously too complex to permit Since the iodine(1) ester decomposes the more rapidly, decomposition of ethyl iodide could be ac- easy kinetic measurements of the type that could celerated during some stages of the reaction. It is eliminate some of these alternatives. We are even possible to imagine the reactants in equation 7 therefore abandoning further study of these reactions. We would be happy t o make our detailed to form products that led to chain branching. It is less easy to account for the observed in- data available t o anybody interested in further creases in the apparent HI08 measurements. It study of these peculiar reactions.
+
+
+
2981
A4r,a.i~or~s OF Geissospermum vellosii
SEPTEMBER, 1962
Experimental Materials.-Dry silver propionate was prepared from the acid by standard procedures. Resublimed reagent grade iodine was used without further purification. Solutions could be labeled with carrier-free obtained from the Oak Ridge Sational Laboratory. The saturated hydrocarbon solvent was mostly hexane. It was prepared from Skellysolve B by treatment with fuming sulfuric acid. Tests with 1'31 indicated that different preparations contained impurities capable of reacting with no more than 6 X 10" niole/l. of iodine. Procedure.-Because of the extreme sensitivity to moisture of iodine(II1) tripropionate, solutions were dried over phosphorus pentoxide and all experiments were conducted in an atmosphere of dry carbon dioxide. Samples of solid silver propionate were treated with hexane solutions of iodine, and the resulting silver iodide was removed by filtration during transfer of the solution to a thermostated flask that was kept in the dark under dry carbon dioxide. Aliquots could be withdrawn a t appropriate times for analysis. Analytical Procedures.-Concentrations of were determined with thin-walled Geiger tubes. Empirical corrections were applied because of different efficiencies for counting iodine in water and in hexane solution. Counting procedures could be used to determine number of moles of iodine in a particular phase regardless of its chemical form.
Elementary iodine in hexane was determined with a Beckman DU spectrophotometer. Ethyl iodide was also determined spectrophotometrically. The molar extinction coefficient of 451 zk 5 at 259 mb agreed well with the value reported by Seheibe.8 The preEence of this compound in a reacted mixture was also confirmed by forming the ethyl 3,5-&nitrobemoate and taking a mixed melting point with the derivative prepared from commercial ethyl iodide. As has been indicated above, positive iodine esters were determined by shaking the hexane solution with water, adding an excess of acid and iodide to the aqueous layer, and titrating with thiosulfate. In some experiments in which the hexane layer contained only iodine(II1) ester, the hexane layer was also reduced with aqueous sulfitc and titrated with silver nitrate. The combined argentometric and iodometric procedures demonstrated that iodate was the only iodine containing species present in the aqueous extract.
Acknowledgment.-This
work was supported
in part by the U.S.Atomic Energy Commission under Contract AT(45-1)-1310. (8) G . Scheibe, Ber., 68, 592 (192.5).
Alkaloids of Geissosperrnum uellosii. Isolation and Structure Determinations of Vellosimine, Vellosiminol, and Geissolosimine HENRYRAPOPORT AND RICHARD E. MOORE' Department of Chemistrv, University of California, Berkeley, California Received April 6, 1962 By systematic fractionation of the alkaloid-rich pH 7 fraction* from G. velloszi three new alkaloids, viz., vellosirnine, vellosiminol, and geissolosimine,have been isolated. Vellosomhe and vellosiminol are indoles, and by n.m.r. studies and comparison with known alkaloids, have been shown to have structures I and 111,respectively. Geissolosimine (formerly called a 2 ) is a dimeric alkaloid and i s cleaved by acid to the indole, vellosimine, and the indoline, geissoschieoline. From this cleavage reaction, its reconstitution from velloeimine and geissoschizcline, and its n.m.r. spectrum, geissolosimine lras been assigned structure VII.
A detailed procedure has been described2 for the separation of the Geivsospermum vellosii alkaloids into various fractions by systematic liquid-liquid extraction a t a series of differeiit pH's. From the alkaloid-rich pH 7 fraction, geissospermine was isolated, usually by crystallization, and large-scale chromatography on alumina then was applied to the residual material. I n this way, a number of ohher alkaloids were isolated, the chief one being alkaloid D2. By further refinement of the pH 7 fraction, we have isolated two additional crystalline alkaloids. These new alkaloids, vellosimine and vellosiminol, and geissolosimine (alkaloid D2) are the subject of the present report. Isolation.-The large amounts of material accumulated after removing geissospermine from the pH 7 fraction made detailed alumina chrornatography quite unappealing as a fractionation proce(1) National Institutes of Health Predoctoral Fellow. (2) H. RapoDorC, T.P. Onak, N. A. Hughes, and M. G. Reinecke, J . Am. C h r m SOC.,80,1801 (1958).
dure. Instead, a pass through a short alumina column first was applied and this clearly removed about one fourth of the material as crystalline geissoschizoline.2 The remaining three fourths was subjected to systematic, continuous liquidliquid extraction a t pH 6, 6.6, and 10. From the pH 10 fraction, more geissoschizoline was isolated, and the pH 6.6 fraction yielded additional geissospermine. The pH 6 fraction, originally obtained by extraction with ether, now was separated into two portions by extraction with benzene a t pH 4.9, fraction 1, and ether a t pH 7, fraction 2. Further purification of fraction 2 by chromatography on alumina was fruitful and resulted in the isolation of two crystalline minor alkaloids, v e l b simine and vellosiminol. Geissolosimine was not isolated anew; it was the material (alkaloid D,) obtained by the previous procedure.2 Structure Determinationa-Vellosimine has the molecular formula CleHEoN20and an ultraviolet
